Exploring brain structure differences and male/female classification: a machine learning study on Huffaz and non-Huffaz

Sapuan, Abdul Halim and Abdul Majid, Zafri Azran and Mohd Tamrin, Mohd Izzuddin and Che Azemin, Mohd Zulfaezal (2025) Exploring brain structure differences and male/female classification: a machine learning study on Huffaz and non-Huffaz. In: 2024 IEEE-EMBS Conference on Biomedical Engineering and Sciences (IECBES), Penang, Malaysia.

PDF - Published Version Restricted to Registered users only Download (581kB) | Request a copy

Abstract

This study investigates the potential of machine learning (ML) to classify sex based on brain structural features, incorporating the unique perspective of cognitive training through Quran memorization. Using T1-weighted MRI scans from 47 healthy young adults (23 Huffaz, 24 non-Huffaz), we extracted features related to brain volume and fractal dimensions (box-counting and Fourier). Four ML algorithms (logistic regression, support vector machine, deep learning, random forest) were employed to classify sex, with Huffaz/non- Huffaz status as a predictor. Results showed high accuracy (up to 92.86%) with random forest and deep learning, particularly when combining multiple features. This suggests the potential of brain structure as a sexual biomarker and highlights the impact of cognitive training on brain morphology. Further research with larger, diverse cohorts is needed to validate these findings and explore the specific neural correlates underlying sex differences and the influence of cognitive practices.

Item Type:	Proceeding Paper (Poster)
Uncontrolled Keywords:	sex classification, MRI, machine learning, Huffaz
Subjects:	R Medicine > R Medicine (General) T Technology > T Technology (General)
Kulliyyahs/Centres/Divisions/Institutes (Can select more than one option. Press CONTROL button):	Kulliyyah of Allied Health Sciences
Depositing User:	Dr. Mohd Zulfaezal Che Azemin
Date Deposited:	20 May 2025 16:24
Last Modified:	20 May 2025 16:24
URI:	http://irep.iium.edu.my/id/eprint/121140

Actions (login required)

View Item